But I think there is a message here - badly phrased perhaps - that
running code is needed for such proposals to be thoroughly considered.
Suppose there was a proposal that all RFCs should be sourced as XML
files. We have a lot of running code to measure that proposal against.
Douglas Otis started work on running code for his alternative to ISDs.
I'm sure that's the best way to bring ideas in this space forward.
In general I agree that running code is the strongest indicator - one
way or another. This having been said, process experiments with the
document series are troublesome, because at the end of the experiment we
are left in a situation where we have to clean it up, usually by further
annotating the documents in some way that hopefully does not go against
the intentions of prior decisions. Tricky stuff.
I am in favor, instead, of an evolution as opposed to revolution. Let
me stop there. I'll have a draft in the coming days over which we can
Ietf mailing list